Discrimination and Calibration of the Estimated Post-Transplant Survival Model by Race
Bibliographic record
Abstract
Background: The Estimated Post-Transplant Survival (EPTS) score is widely used to predict survival in kidney transplant recipients and guide high-longevity kidney allocation. Prior studies identified racial disparities in access to high-longevity kidneys under this system, but EPTS performance across racial groups has not been directly evaluated. We evaluated EPTS model performance across racial groups. Methods: This retrospective cohort study included first-time adult deceased donor kidney transplant recipients from the U.S. Scientific Registry of Transplant Recipients (2013–2023). Cox proportional hazards models estimated mortality using Raw EPTS, with an interaction term for race (White, Black, Other). Hazard ratios (HRs) per unit Raw EPTS were calculated for each group. Discrimination was assessed using overall and time-dependent Harrell’s C-statistic, compared using DeLong’s test. Calibration was evaluated at 1, 3, and 5 years using calibration plots. Results: Among 123,952 recipients, 67,027 (54%) were White, 44,471 (36%) were Black, and 12,454 (10%) were Other. A significant interaction between Raw EPTS and race was detected (p<0.001). The HR per unit Raw EPTS was higher for White recipients (HR 3.66; 95% CI: 3.53, 3.80) than Black recipients (HR 2.91; 95% CI: 2.80, 3.02). Discrimination was slightly lower for Black recipients (C-statistic 0.678) than White recipients (0.712, p<0.001). Re-fitting the EPTS coefficients within each race group did not improve discrimination. Calibration was best at 1 year, declining at 3 and 5 years, but remained similar across groups. Conclusion: The association between Raw EPTS and mortality differed by race, with a stronger impact in White recipients. Discrimination was modestly lower for Black recipients, reducing EPTS’s ability to rank risk. Despite this, calibration was similar, providing accurate 1-year survival predictions across groups. These findings support continued EPTS use but highlight the need to monitor model performance, ensure fairness, and explore ways to improve baseline risk-stratification for Black recipients.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.083 | 0.139 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.003 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.002 | 0.003 |
| Research integrity | 0.002 | 0.003 |
| Insufficient payload (model declined to judge) | 0.004 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".